We built this because the classroom hadn't changed in a hundred years.

The problem

Walk into almost any school today and watch what happens. Thirty children sit in rows. One curriculum moves at one pace. The kid who already understands fractions stares at the ceiling. The kid who doesn't yet — too embarrassed to ask the question for the third time — quietly falls a year behind.

It is, when you stop to look at it, an absurd design. We've built personalization into our music, our shopping, our news feeds. But the system we use to shape children's minds is still, in essence, a 19th-century factory: same input, same output, same speed, regardless of the human inside.

For the kids on the edges — gifted, struggling, anxious, neurodivergent, simply different — this isn't a small inefficiency. It's the difference between a child who loves school and a child who quietly decides, around age nine, that they're not a learner.

The moment we realized the math had changed

We had been thinking about this for years before we actually did anything about it. Because true self-paced learning — the kind where every child has their own teacher, working at their own level, on their own thing — has always been a known good. It's how every wealthy family with means has educated their children for centuries: tutors.

But it doesn't scale. You cannot put a private tutor next to every child in every classroom. The math doesn't work. Not on a school's budget. Not on most parents' budgets.

Suddenly the price of an attentive, patient, infinitely-knowledgeable tutor wasn't sixty dollars an hour. It was something close to free. The bottleneck on personalized learning had never really been pedagogy — we've known how to teach individual children well for a long time. The bottleneck was always cost. AI removed it.

Why we started with schools, not parents

The obvious move would have been to build a consumer app. Build a tutor app, sell it to parents, ride the wave. We thought hard about this and didn't do it. Three reasons.

First, the children who most need personalization — the ones for whom the system is actively failing — are the ones whose parents are least likely to download a $30/month tutor app. Going direct-to-parent serves the families who are already winning. We wanted to serve the kid in the back of the classroom whose parents are working two jobs.

Second, school is where children spend most of their waking hours. Anything that doesn't change school is, at best, a workaround. We wanted to change the thing itself.

Third — and this is the founder-economics reason — schools have a budget for what we sell. Parents have a budget for things called "essentials" and a much smaller budget for things called "extras." Personalized education will, over the next decade, become an essential. But the business that bridges the gap between today and that future has to be paid for by institutions, not impulse purchases.

The product, in two acts

So we sat down with twenty principals — large schools, small schools, urban, rural, secular, religious — and asked them what was actually broken. They were extraordinarily generous with their time. We learned that there are really two products hiding inside the same idea, and they need to be sold differently.

Act One is AI Copilot. It's the version of Pace that doesn't ask the school to change anything. The bell rings, the classes happen, the teachers teach. But underneath, every student is on a personalized path. Mastered the topic? Skip it. Stuck? Get a different scaffold. Parents see what's actually happening for the first time. Schools see modest revenue uplift. Teachers feel relief, not threat.

Act Two is Full Self-Paced. This is the version where the school commits to changing the underlying model — students moving between subject levels independently, continuous intake, the possibility of graduating early. The upside for families is enormous. The upside for schools is enormous. But it requires conviction from the school's leadership, and it requires the kind of marketing support most schools have never had to do. So we provide it.

What we got wrong (and fixed)

Two things.

First, we initially priced like a consumer SaaS — a per-student monthly fee. Schools hated it. Not because of the dollar amount, but because it felt like another IT line-item rather than a bet on growth. So we restructured: for school-wide deployment, our fee is a small percentage of the new tuition uplift we help unlock. We win when they win. We don't if they don't.

Second, we tried to ship a beautiful all-in-one platform, all features, day one. It scared schools. Now we ship Act One first — the gentler, plug-and-play version — and earn the right to introduce Act Two once the school has seen the results.

Where we're going

The honest answer is: we don't know exactly. We know what we're sure of and what we're still figuring out.

What we're sure of: in ten years, the idea of thirty children moving in lockstep through identical material will look as strange as the idea of bloodletting. Children will learn at the speed and depth that's right for them. The teacher's job will change from front-of-the-room performer to coach, mentor, and connector. Parents will know, in real time, what their child needs and what they're getting.

What we're still figuring out: which subjects are easiest to truly self-pace, what role AI should play in the affective dimension of learning (motivation, persistence, identity), how schools redistribute teacher time when classes aren't synchronous, and what credentialing looks like when grade levels stop being meaningful.

We have a decade or two of work ahead. We'd like to do it with you.

The roadmap, briefly

1. 2024 Co-design with 20 principals across 6 countries. First pilot in two schools, 130 students total.
2. 2025 AI Copilot generally available. Founding-school program launched. ~1,200 students on platform.
3. 2026 · Now Full Self-Paced mode in early access. Three randomized trials in progress. First grade-agnostic cohort graduates next term.
4. 2027 Public peer-reviewed publication of two-year mastery and equity outcomes. Expansion into secondary.
5. 2028+ Independent accreditation pathway for grade-agnostic schools. Open API for curriculum partners.